Determination device and determination method

Abstract

To provide a determination device and a determination method with which it is possible to determine the proficiency level of receiving staff with high accuracy.SOLUTION: The determination device comprises: a proficiency level acquisition unit that acquires the proficiency level information in which the proficiency level of receiving staff remotely operating, by a monitoring board, an automatic toll collection machine that receives tolls from vehicles is evaluated for each work item; a data extraction unit that extracts, for each work item, system data pertaining to the operation of receiving staff from the logs recorded in either the automatic toll collection machine or the monitoring board; a teaching data creation unit that creates teaching data from the system data and the proficiency level information; a learning model creation unit that creates from the teaching data a learning model that has a correlation between the system data and the proficiency level information; and a proficiency level determination unit that determines the proficiency level of receiving staff for each work item by inputting, to the learning model, the system data extracted, apart from the creation of teaching data, from the logs recorded in either the automatic toll collection machine or the monitoring board.SELECTED DRAWING: Figure 3

Description

Technical Field

[0001] The present disclosure relates to a determination device and a determination method.

Background Art

[0002] For example, at a tollgate on a toll road, an automatic toll collector for collecting tolls is installed. Some automatic toll collectors are remotely operated by toll collectors. When collecting tolls using such an automatic toll collector, differences in service occur depending on the proficiency of the toll collector. Therefore, it is necessary to provide education to toll collectors with low proficiency. In order to improve the efficiency of education for toll collectors, it is preferable to accurately determine the proficiency. For example, Patent Document 1 below discloses a determination device that determines the proficiency of a task performer based on the situation of the task and the mental state of the task performer.

Prior Art Documents

Patent Documents

[0003]

Patent Document 1

Summary of the Invention

Problems to be Solved by the Invention

[0004] However, in the determination device of Patent Document 1, there are cases where the situation of the task itself cannot be objectively evaluated, so the determination of proficiency cannot be sufficiently objective. In particular, in the field of toll collection on toll roads, a method for objectively determining proficiency has not been established, and it has been difficult to accurately determine proficiency.

[0005] The present disclosure has been made in view of the above situation, and an object thereof is to provide a determination device and a determination method capable of highly accurately determining the proficiency of a toll collector.

Means for Solving the Problems

[0006] To solve the above problems, the determination device according to this disclosure includes: a proficiency acquisition unit that acquires proficiency information evaluated for each work item of the proficiency of a toll collector who remotely operates an automatic toll collection machine for collecting tolls from vehicles using a monitoring panel; a data extraction unit that extracts system data related to the toll collector's operation for each work item from logs recorded in at least one of the automatic toll collection machine and the monitoring panel; a teacher data creation unit that creates teacher data from the system data and the proficiency information; a learning model creation unit that creates a learning model having a correlation between the system data and the proficiency information from the teacher data; and a proficiency determination unit that inputs the system data extracted from logs recorded in at least one of the automatic toll collection machine and the monitoring panel, separately from the use for creating the teacher data, into the learning model to determine the proficiency of the toll collector for each work item.

[0007] The determination method relating to this disclosure is: Computer The process involves obtaining proficiency information, which evaluates the proficiency of toll collectors who remotely operate an automated toll collection machine that collects tolls from vehicles using a monitoring panel, for each work item. The aforementioned computer The steps include: extracting system data related to the operator's operations for each work item from logs recorded in at least one of the automatic toll collection machine and the monitoring panel; The aforementioned computer A step of creating training data from the aforementioned system data and the aforementioned proficiency information, The aforementioned computer The steps include creating a learning model from the aforementioned training data that has a correlation between the system data and the proficiency information, The aforementioned computer The learning model includes, separately from the use of creating the training data, the logs recorded in at least one of the automatic toll collection machine and the monitoring panel. Ta The system includes the step of inputting the aforementioned system data and determining the proficiency level of the recipient for each work item. [Effects of the Invention]

[0008] According to the determination device and determination method of this disclosure, the proficiency level of the recipient can be determined with high accuracy. [Brief explanation of the drawing]

[0009] [Figure 1] This figure shows the overall configuration of the fee collection system according to the embodiment of this disclosure. [Figure 2] This figure shows the functional configuration of the fee collection system according to the embodiment of this disclosure. [Figure 3] This figure shows the functional configuration of the determination device according to the present disclosure. [Figure 4] This flowchart shows the procedure for the determination method according to the embodiment of this disclosure. [Figure 5] This flowchart shows the procedure for the training data creation phase of the determination method according to the embodiment of this disclosure. [Figure 6] This is a table showing an example of proficiency information according to the embodiments of this disclosure. [Figure 7] This is a table showing an example of system data for a certain recipient according to an embodiment of this disclosure. [Figure 8] This flowchart shows the steps of the determination phase of the determination method according to the embodiment of this disclosure. [Figure 9] A schematic block diagram showing the configuration of a computer according to one embodiment of this disclosure. [Figure 10] This figure shows the functional configuration of the determination device relating to a modified example of this disclosure. [Figure 11] This flowchart shows the procedure for the training data creation phase of the determination method relating to the modified example of this disclosure. [Figure 12] This flowchart shows the procedure for the determination phase of the determination method relating to a modified example of this disclosure. [Modes for carrying out the invention]

[0010] The determination device 5 and determination method according to the embodiments of this disclosure will be described below with reference to Figures 1 to 8. The determination device 5 is introduced into, for example, the toll collection system 1 shown in FIG. 1. The determination device 5 determines the proficiency level of the toll collectors in the toll collection system 1.

[0011] (Configuration of the toll collection system) FIG. 1 is a diagram showing the overall configuration of the toll collection system 1 according to an embodiment of the present disclosure. The toll collection system 1 according to the present embodiment is provided, for example, at a toll gate (entrance toll gate or exit toll gate) of a toll road. The toll collection system 1 collects tolls from vehicles traveling on the lane L connecting the toll road and the general road. In the present embodiment, the toll collection system 1 can automatically collect tolls from a vehicle (hereinafter also referred to as "ETC vehicle A1") equipped with an in-vehicle unit α using an electronic toll collection system (ETC: Electronic Toll Collection System (registered trademark), also referred to as "automatic toll collection system"). The lane L in the present embodiment is a mixed lane on which both ETC vehicle A1 and non-ETC vehicle A2 travel, and the toll collection system 1 collects tolls from both ETC vehicle A1 and non-ETC vehicle A2.

[0012] In the present embodiment, an example in which the toll collection system 1 is provided at an exit toll gate will be described. Hereinafter, the direction in which the lane L extends (±X direction) will be referred to as the "lane direction". Also, the toll road side (-X side) in the lane direction will be referred to as the "upstream side", and the general road side (+X side) will be referred to as the "downstream side". Further, the width direction (±Y direction) of the lane L will be referred to as the "lane width direction".

[0013] In addition, although FIG. 1 shows an example in which only one lane L is provided at the toll gate, the present invention is not limited to this. In other embodiments, a plurality of lanes L may be provided at the toll gate.

[0014] As shown in FIG. 1, the toll collection system 1 includes an ETC facility 2, an automatic toll collector 3, and a monitoring panel 4.

[0015] (Configuration of the ETC facility) ETC equipment 2 is a group of devices for performing a first toll collection process, which automatically collects tolls via wireless communication from an ETC vehicle A1 equipped with an on-board unit α. Specifically, as shown in Figure 1, ETC equipment 2 includes an entry-side vehicle detector 21, a roadside antenna 22, a communication vehicle detector 23, a lane server 24, a start control unit 25, and an exit-side vehicle detector 26.

[0016] The entry-side vehicle detector 21 is installed on islands I provided on both sides of lane L at the entry detection position X1 on the upstream side (-X side) in the lane direction, and detects the entry of a vehicle into lane L. The entry-side vehicle detector 21 is a through-type vehicle detector, as shown in Figure 1, for example, and has a pair of light emitters and light receivers installed opposite each other on either side of lane L. The entry-side vehicle detector 21 outputs a detection signal that can distinguish whether or not a vehicle is present at the entry detection position X1 based on whether or not the light receiver has received light emitted from the light emitter. While the light receiver is receiving light, the entry-side vehicle detector 21 outputs a detection signal indicating that no vehicle is present at the entry detection position X1. On the other hand, while the light receiver is not receiving light, the entry-side vehicle detector 21 outputs a detection signal indicating that a vehicle is present at the entry detection position X1.

[0017] The roadside antenna 22 is installed above the road surface downstream (+X side) in the lane direction from the entering vehicle detector 21. When the entering vehicle detector 21 detects that vehicle A has entered lane L, the roadside antenna 22 emits radio waves for wireless communication toward a predetermined communication area R1 on lane L. The communication area R1 is set to the range from the entering vehicle detector 21 to the communication vehicle detector 23, for example, as shown in Figure 1. The roadside antenna 22 transmits and receives various commands and information for executing the first toll collection process with the on-board unit α of ETC vehicle A1 located within the communication area R1 via wireless communication.

[0018] The communication vehicle detector 23 is installed on island I at the communication termination position X2 downstream (+X side) in the lane direction from the roadside antenna 22, and detects when a vehicle passes the communication termination position X2. The communication vehicle detector 23 has the same equipment configuration as the entry vehicle detector 21. When the communication vehicle detector 23 detects that a vehicle has passed the communication termination position X2, the roadside antenna 22 stops emitting radio waves.

[0019] The lane server 24 performs a first toll collection process to collect tolls from ETC vehicle A1. For example, the lane server 24 obtains information from the on-board unit α via the roadside antenna 22, including the vehicle type of ETC vehicle A1, the location (entrance toll gate) and date and time when ETC vehicle A1 entered the toll road, and the payment method for the toll (credit card number, etc.). Based on this information, it calculates the toll according to the vehicle type and distance traveled by ETC vehicle A1, and performs processing such as settlement. The lane server 24 may be installed on island I as shown in Figure 1, or it may be installed inside the toll gate office.

[0020] The departure control unit 25 is installed on island I downstream (+X side) in the lane direction from the toll collection machine 3. The departure control unit 25 raises and lowers the opening and closing bar according to the opening and closing commands from the lane server 24 and the toll collection machine 3, thereby opening (permitting vehicle passage) and closing (restricting vehicle passage) lane L. When the lane server 24 completes the first toll collection process, it outputs an open command to the departure control unit 25 to permit vehicle passage. Also, when the lane server 24 detects that a vehicle has exited lane L using the exit vehicle detector 26 (described later), it outputs a close command to the departure control unit 25 to restrict the passage of following vehicles. Similarly, when the toll collection machine 3 collects a toll from a non-ETC vehicle A2 in the second toll collection process, it outputs an open command to the departure control unit 25 to permit vehicle passage. Furthermore, when the automatic toll collection machine 3 receives an instruction to complete the second toll collection process, it outputs a closing command to the departure control unit 25, thereby restricting the passage of following vehicles.

[0021] The exit-side vehicle detector 26 is installed on island I at the exit detection position furthest downstream (+X side) in the lane direction, and detects when a vehicle passes the exit detection position (exits from lane L). The exit-side vehicle detector 26 has the same equipment configuration as the entry-side vehicle detector 21.

[0022] Although Figure 1 shows an example where the entry vehicle detector 21, the communication vehicle detector 23, and the exit vehicle detector 26 are through-type vehicle detectors, the system is not limited to this. In other embodiments, these may be reflective vehicle detectors.

[0023] Furthermore, ETC equipment 2 may utilize existing equipment already installed in conventional ETC-only lanes or mixed-use lanes.

[0024] (Configuration of an automated fare collection machine) The automated toll collection machine 3 executes a second toll collection process to collect tolls from non-eligible vehicles that could not be collected by the first toll collection process performed by the ETC equipment 2. Non-eligible vehicles include non-ETC vehicles A2 that do not have an on-board unit α installed. Furthermore, ETC vehicle A1 includes normal ETC vehicle A1a that was able to correctly collect tolls by the first toll collection process, and abnormal ETC vehicle A1b that was unable to collect tolls by the first toll collection process. Therefore, abnormal ETC vehicle A1b is included in the non-eligible vehicles. Possible causes for the first toll collection process for ETC vehicle A1 not completing successfully include malfunctions of ETC equipment 2 or on-board unit α, communication errors, and expired credit cards inserted in on-board unit α. The automated toll collection machine 3 is connected to the lane server 24 of ETC equipment 2 in a way that allows for sending and receiving data. For this reason, the automated toll collection machine 3 can transmit logs from ETC equipment 2 to an external source. In this embodiment, the example described is one in which the non-target vehicle is a non-ETC vehicle A2.

[0025] As shown in Figure 1, the automatic toll collection machine 3 is installed on island I in the lane direction, downstream of the communication vehicle detector 23 (+X side) and upstream of the start control unit 25 (-X side). For example, the automatic toll collection machine 3 is positioned so as to coincide with the position of the driver's seat of a vehicle that stops in front of the start control unit 25 (-X side) when the opening / closing bar of the start control unit 25 is closed.

[0026] Figure 2 shows the functional configuration of the toll collection system 1 according to the embodiment of this disclosure. As shown in Figure 2, the automatic toll collection machine 3 includes a camera 31, an operation reception unit 32, and a toll collection processing unit 33. The camera 31 and the operation reception unit 32 are located on the side of the automatic toll collection machine 3 facing the lane L side (the +Y side in the example in Figure 1), as shown in Figure 1.

[0027] Camera 31 functions as a notification unit that detects when a non-ETC vehicle A2 enters lane L and when it exits lane L, and notifies the monitoring panel 4. Camera 31 photographs the shooting area R2, which includes the installation location of the automatic toll collection machine 3, and by applying predetermined image processing to the captured video (image), it detects the presence or absence of a vehicle in the shooting area R2 and whether or not the vehicle is moving. As a result, camera 31 detects when a non-ETC vehicle A2 is parked at the installation location of the automatic toll collection machine 3 (entering lane L), and when the parked non-ETC vehicle A2 moves downstream (towards +X) (exiting lane L). Note that camera 31 may be a conventional surveillance camera installed on the automatic toll collection machine 3.

[0028] The operation reception unit 32 accepts operations and toll payments from passengers of non-ETC vehicle A2. For example, the operation reception unit 32 has a ticket slot for inserting toll tickets, a cash slot for accepting cash payments, a change return slot for dispensing change, a card slot for inserting IC cards (credit cards, etc.), and an intercom for calling a toll collector.

[0029] The toll collection processing unit 33 executes a second toll collection process to collect tolls from non-ETC vehicle A2 (a vehicle not subject to the first toll collection process) from which tolls could not be collected in the first toll collection process. The toll collection processing unit 33 starts and ends the second toll collection process according to instructions from the monitoring panel 4. When the toll collection processing unit 33 has collected tolls from non-ETC vehicle A2, it outputs an open command to the departure control unit 25 to allow the vehicle to pass. When the automatic toll collection machine 3 receives an instruction from the monitoring panel 4 to end the second toll collection process, it outputs a close command to the departure control unit 25 to restrict the passage of following vehicles.

[0030] (Configuration of the monitoring panel) The monitoring panel 4 is installed in a remote location such as a toll booth office and is connected to the automatic toll collection machine 3 via a communication network such as the internet. The monitoring panel 4 is used by toll collectors stationed at the toll booth office to remotely monitor and operate the automatic toll collection machine 3.

[0031] As shown in Figures 1 and 2, the monitoring panel 4 includes a display device 41, an input device 42, and an instruction transmission unit 43.

[0032] The display device 41 displays various data received from the toll collection machine 3. For example, when the toll collection machine 3 notifies the display device 41 of the entry of a non-ETC vehicle A2, the display device 41 displays the video captured by the camera 31 of the toll collection machine 3. The display device 41 may also display various information acquired and generated by the toll collection machine 3 during the second toll collection process (information read from the toll ticket, calculated toll charges, etc.).

[0033] The input device 42 accepts input operations from toll collectors stationed at the toll booth office. The input device 42 may also be a touch panel on the display device 41. When the toll collection machine 3 notifies the input device 42 that a non-ETC vehicle A2 has entered lane L, the input device 42 accepts operations from the toll collector to instruct the start and end of the second toll collection process.

[0034] The input device 42 is equipped with buttons for use in the event of an abnormal situation, such as a wrong entry button, a non-payment button, and a force-through button. The wrong entry button is pressed by the toll collector when a vehicle mistakenly enters the toll collection system 1. The non-payment button is pressed when the correct toll is not paid to the toll collection system 1. The force-through button is pressed when a vehicle attempts to force its way through the toll collection system 1 without paying the toll, even though it is aware of the toll obligation. In addition to the buttons mentioned above, the input device 42 is also equipped with, for example, a correction button to correct errors made by the payer, and a clear button to cancel the payer's operation.

[0035] The instruction transmission unit 43 transmits the start instruction and end instruction for the second fare collection process, which are input via the input device 42, to the fare collection machine 3.

[0036] The fare collection system 1 described above results in differences in the services provided to users depending on the fare collector's skill level. Therefore, it is necessary to provide training to fare collectors with low skill levels. To select fare collectors who require training, a skill level assessment device 5 has been introduced.

[0037] (Configuration of the determination device) Figure 3 shows the functional configuration of the determination device 5 according to the present disclosure. The determination device 5 is connected to the automatic fare collection machine 3, for example, by wireless communication such as a network. As shown in Figure 3, the determination device 5 comprises the following functional units: a proficiency acquisition unit 51, a data extraction unit 52, a first filtering unit 53, a second filtering unit 54, a third filtering unit 55, a teacher data creation unit 56, a learning model creation unit 57, and a proficiency determination unit 58.

[0038] The proficiency acquisition unit 51 receives and acquires proficiency information, which evaluates the proficiency of the recipient for each work item. In this embodiment, the proficiency acquisition unit 51 receives the proficiency information as a score. The data extraction unit 52 extracts system data related to the toll collector's operations for each work item from the logs recorded in both the automatic toll collection machine 3 and the monitoring panel 4. In this embodiment, the data extraction unit 52 extracts the system data after filtering by the first filtering unit 53, the second filtering unit 54, and the third filtering unit 55, which will be described later.

[0039] The first filtering unit 53 performs filtering based on the recipient's ID included in the log, excluding logs related to recipients for whom proficiency assessment is not required. The second filtering unit 54 performs filtering to exclude logs related to vehicles other than those that require operation by a receipt-person.

[0040] The third filtering unit 55 removes logs when an inappropriate operation is performed by the recipient for determining proficiency level.

[0041] The teacher data creation unit 56 creates teacher data from system data and proficiency information. The learning model creation unit 57 uses system data and proficiency information as training data to create a learning model that has a correlation between system data and proficiency information. The proficiency determination unit 58 inputs system data extracted from logs recorded in at least one of the automatic fare collection machine 3 and the monitoring panel 4, separate from the data used to create the teacher data, into the learning model to determine the proficiency of the fare collector for each work item. In this embodiment, the proficiency determination unit 58 determines the proficiency of the fare collector using a score.

[0042] (Procedure for determination) Next, we will explain the method for determining the proficiency level of the toll collection officer using the determination device 5.

[0043] Figure 4 is a flowchart showing the procedure of the determination method according to the embodiment of this disclosure. The determination method of this embodiment includes a training data creation phase S1, a learning phase S2, and a determination phase S3, as shown in the processing flow of Figure 4. First, the training data creation phase S1 is executed.

[0044] (Training data creation phase) Figure 5 is a flowchart showing the procedure for the training data creation phase S1 of the determination method according to the embodiment of this disclosure. As shown in the processing flow in Figure 5, in the teacher data creation phase S1, the proficiency acquisition unit 51 first receives and acquires proficiency information in which the proficiency of the recipient is evaluated for each work item (step S11). The proficiency information is information that the instructor has manually determined for the proficiency of each recipient. Instructors monitor the daily work of the collection officers and evaluate their proficiency from the perspective of whether or not further training by the instructors is necessary.

[0045] Figure 6 is a table showing an example of proficiency information according to the embodiment of this disclosure. As shown in Figure 6, the proficiency information includes the level of proficiency for each work item for each collection officer. In the example shown in Figure 6, the work items listed include "proficiency in machine operation (collection)", "proficiency in machine operation (maintenance)", "customer service", and "adherence to collection procedures".

[0046] "Proficiency in machine operation (collection)" refers to the level of proficiency a toll collector has in the operations they perform when actually collecting tolls from vehicles. "Proficiency in machine operation (maintenance)" refers to the level of proficiency in operations related to the maintenance of the automatic fare collection machine 3, such as inspection and maintenance of the automatic fare collection machine 3. "Customer Service" refers to the service provided by fare collectors to passengers on the vehicle. "Adherence to collection procedures" refers to whether the collector performed the operation according to the prescribed procedures. The work items listed above as being subject to evaluation are merely examples. Other work items may also be included as subjects of evaluation.

[0047] Furthermore, in the example shown in Figure 6, the proficiency level of the recipient is determined by a score. This score only needs to be able to distinguish whether or not training is necessary as a result of the instructor's assessment. The criteria for the instructor to determine whether or not training is necessary are arbitrary criteria set by the instructor. However, it is preferable that the assessment criteria be standardized for each business operator using the assessment device 5 and assessment method of this disclosure. Also, scoring is performed for each work item.

[0048] In the example shown in Figure 6, a score of 3 indicates "No further education needed (no further improvement in proficiency is expected)." A score of 2 indicates "Education is desirable (acceptance is not a problem, but proficiency can be improved through education)." A score of 1 indicates "Education is necessary (acceptance is problematic, and improvement in proficiency is necessary through education)." Furthermore, proficiency information is not limited to that evaluated manually by instructors; for example, it may also be evaluated by devices other than the determination device 5 described in this disclosure.

[0049] After step S11, the data extraction unit 52 extracts system data from logs recorded in at least one of the automatic toll collection machine 3 and the monitoring panel 4. In this embodiment, the data extraction unit 52 extracts system data from logs recorded in the automatic toll collection machine 3 and logs recorded in the monitoring panel 4.

[0050] The logs used here are records accumulated in the toll collection system 1 (more specifically, the automatic toll collection machine 3 and the monitoring panel 4) when each toll collector, as evaluated by the instructor, performs their toll collection duties. In this embodiment, system data extraction is performed after a three-stage filtering process.

[0051] First, the first filtering unit 53 filters the logs to exclude logs of receivers for whom proficiency assessment is not required, based on the receiver's ID included in the log (step S12). The receiver's ID used in step S12 is, for example, entered by the receiver into the monitoring panel 4 at the start of their shift. In step S12, for example, logs of highly skilled receivers are excluded. The following criteria for determining whether a receiver is skilled can be arbitrarily set.

[0052] After step S12, the second filtering unit 54 performs filtering to exclude logs related to vehicles other than those that require operation by a toll collector (step S13). Here, the vehicles that require operation are the abnormal ETC vehicle A1b, the non-ETC vehicle A2 that made an intercom call, and the normal ETC vehicle A1a that made an intercom call. In other words, in step S13, for example, only logs related to the normal ETC vehicle A1a that did not call a toll collector via intercom are excluded. This allows data extraction to be performed in the steps described later for logs related to the abnormal ETC vehicle A1b, the non-ETC vehicle A2 that made an intercom call, and the normal ETC vehicle A1a that made an intercom call.

[0053] After step S13, the third filtering unit 55 performs filtering to remove logs when an inappropriate operation is performed by the receiver for determining proficiency (step S14). In step S14, for example, logs are removed when various buttons on the monitoring panel 4 that are pressed in the event of an abnormality are pressed. More specifically, logs are removed when, for example, the incorrect entry button is pressed, the unpaid button is pressed, or the forced entry button is pressed.

[0054] Furthermore, in step S14, logs are excluded if, for example, there is a long time between the vehicle's entry and the intercom call to the toll collector. More specifically, logs are excluded if the time from the entry notification by the entry vehicle detector 21 to the call notification by the automatic toll collection machine 3 is longer than a predetermined time.

[0055] Furthermore, in step S14, logs related to receipts when, for example, equipment malfunctions occur are excluded. More specifically, logs when an abnormality of the automatic fare collection machine 3 is notified are excluded.

[0056] After the above-mentioned filtering is completed (after steps S12 to S14 are completed), the data extraction unit 52 extracts system data related to the toll collector's operations for each work item from the logs recorded in the automatic toll collection machine 3 and the logs recorded in the monitoring panel 4 (step S15). In step S15, for example, "communication logs," "monitoring panel operation logs," and "monitoring panel operation logs" are extracted from the logs as system data. The "communication log" is a record of communication between the automated toll collection machine 3 and the monitoring panel 4. The "monitoring panel operation log" and the "monitoring panel operation log" are both logs related to monitoring panel 4. The "monitoring panel operation log" is mainly a log of when the toll collector operates other devices such as the automatic toll collection machine 3 via monitoring panel 4, while the "monitoring panel operation log" is mainly a log of when monitoring panel 4 itself operates.

[0057] Figure 7 is a table showing an example of system data according to the embodiment of this disclosure. In the example shown in Figure 7, the extracted data is compiled according to the work items shown in Figure 6. Furthermore, supplementary information is entered for the extracted data as needed. In the example shown in Figure 7, for example, data such as "communication log (time from start to finish of receiving)" and "monitoring panel operation log (number of times the correction button and clear button are pressed)" are extracted to assess proficiency in machine operation (receiving).

[0058] The "communication log (time from start to finish of receipt)" is used as an indicator, for example, to show whether the receipt recipient is receiving the data smoothly, while the "monitoring panel operation log (number of times the correction button and clear button are pressed)" is used as an indicator to show the accuracy of the receipt recipient's operation. The system data shown in Figure 7 is merely an example, and other extracted data may be included.

[0059] The filtering by the first filtering unit 53, the second filtering unit 54, and the third filtering unit 55 described above (steps S12 to S14), and the data extraction (step S15) are automatically performed, for example, by a program incorporated into the determination device 5.

[0060] After step S15, the teacher data creation unit 56 creates teacher data from the system data and proficiency information (step S16). In step S16, for example, based on the recipient's ID, the system data and proficiency information collected for the same recipient are linked to create the teacher data. After the above steps, the training data creation phase S1 is completed. Following the training data creation phase S1, the training phase S2 is executed.

[0061] (Learning Phase) In learning phase S2, the learning model creation unit 57 creates a learning model from the training data that has a correlation between system data and proficiency information. The created learning model is, for example, a machine learning model classified as supervised learning. The state information of the learning model is the same system data as the system data collected in training data creation phase S1, and the state information of the learning model is the same proficiency information as the proficiency information collected in training data creation phase S1. Furthermore, the purpose of learning the learning model is to determine the proficiency level of the recipient.

[0062] The learning model can be any machine learning model that serves the purpose of determining the proficiency level of the payer. The type of learning model can be selected as appropriate. After the learning model is created, the judgment phase S3 is executed.

[0063] (Decision Phase) Figure 8 is a flowchart showing the procedure for determination phase S3 of the determination method according to the present disclosure. As shown in Figure 8, in the determination phase S3, the data extraction unit 52 first extracts system data from logs recorded in at least one of the toll collection machine 3 and the monitoring panel 4 for the toll collection officers whose proficiency level is to be determined, separately from the data used to create the training data. In this embodiment, the data extraction unit 52 extracts system data from logs recorded in the toll collection machine 3 and logs recorded in the monitoring panel 4. In this embodiment, system data extraction is performed after a three-stage filtering process. Furthermore, the data extraction in the judgment phase S3 may be performed on a different set of automatic fare collection machines 3 and monitoring panels 4 than those used for data extraction in the training data creation phase S1.

[0064] First, the first filtering unit 53 filters the logs to exclude logs related to recipients for whom proficiency assessment is not required, based on the recipient's ID included in the log (step S31). In step S31, filtering is performed by the first filtering unit 53, similar to step S12 in the training data creation phase S1.

[0065] After step S31, the second filtering unit 54 performs filtering to exclude logs related to vehicles other than those that require operation by a receipt operator (step S32). In step S32, filtering is performed by the second filtering unit 54, similar to step S13 of the training data creation phase S1.

[0066] After step S32, the third filtering unit 55 performs filtering to remove logs indicating that an inappropriate operation was performed by the recipient for determining proficiency (step S33). In step S33, filtering is performed by the third filtering unit 55, similar to step S14 of the training data creation phase S1.

[0067] After the above-mentioned filtering is completed (after steps S31 to S33 are completed), the data extraction unit 52 extracts system data related to the operations of the toll collector subject to judgment from the logs recorded in the automatic toll collection machine 3 and the logs recorded in the monitoring panel 4, for each work item (step S34). In step S34, system data is extracted in the same way as in step S15 of the training data creation phase S1. Therefore, system data similar to that in step S15 is extracted, as shown in Figure 7.

[0068] After step S34, the proficiency determination unit 58 inputs the new system data extracted in step S34 into the learning model and determines the proficiency level of the collection officer for each work item (step S35). In step S35, proficiency information similar to that obtained in step S11 of the teacher data creation phase S1 is output. That is, in this embodiment, the proficiency determination unit 58 determines the proficiency of each receiver with a score. More specifically, as shown in Figure 6 in step S11, for each receiver, the proficiency is determined with a score of 1 to 3 on a three-level scale for each work item.

[0069] (Effects and Benefits) The determination device 5 and determination method with the above configuration provide the following effects.

[0070] In this embodiment, the determination device 5 comprises a proficiency acquisition unit 51, a data extraction unit 52, a teacher data creation unit 56, a learning model creation unit 57, and a proficiency determination unit 58. The proficiency acquisition unit 51 acquires proficiency information, which evaluates the proficiency of toll collectors who remotely operate the toll collection machine 3, which collects tolls from vehicles, via a monitoring panel 4, for each work item. The data extraction unit 52 extracts system data related to the toll collector's operation for each work item from logs recorded in at least one of the toll collection machine 3 and the monitoring panel 4. The teacher data creation unit 56 creates teacher data from the system data and proficiency information. The learning model creation unit 57 creates a learning model from the teacher data that has a correlation between the system data and the proficiency information. The proficiency determination unit 58 inputs system data extracted from logs recorded in at least one of the automatic fare collection machine 3 and the monitoring panel 4, separate from the data used to create the teacher data, into the learning model to determine the proficiency level of the fare collector for each work item.

[0071] As a result, once a learning model is created, the judgment device 5 can use the learning model to automatically determine the proficiency level of the cashier for each work item. Therefore, the proficiency level of the cashier can be determined without differences in evaluation criteria among instructors. Furthermore, by creating training data from as much system data and proficiency information as possible, the accuracy of the learning model is improved. Therefore, the judgment method used by the judgment device 5 becomes objective and highly accurate. Consequently, the proficiency level of the cashier can be determined with high accuracy. In addition, since proficiency is determined automatically, the efficiency of proficiency determination can be greatly improved.

[0072] Furthermore, because the proficiency level of toll collectors can be determined with high accuracy and efficiency, only those toll collectors who require training can be reliably and quickly selected. Therefore, toll collectors who require training can be effectively selected. This eliminates differences in proficiency levels among toll collectors, enabling consistent service to users of the toll collection system 1.

[0073] In this embodiment, the determination device 5 further includes a first filtering unit 53. The first filtering unit 53 filters logs to exclude logs relating to recipients for whom proficiency level determination is not required, based on the recipient's ID included in the log. The data extraction unit 52 extracts system data from the log after filtering by the first filtering unit 53.

[0074] This allows system data to be extracted while excluding logs related to highly skilled veteran receivers for whom proficiency assessment is unnecessary. Therefore, in the training data creation phase S1, system data can be extracted only from the logs of receivers for whom proficiency assessment is required. This improves the accuracy of the learning model created in the subsequent learning phase S2. As a result, the proficiency level of receivers can be determined with even greater accuracy.

[0075] Furthermore, in the judgment phase S3, system data can be extracted only from the logs of the receivers whose proficiency needs to be assessed. Therefore, proficiency assessment can be performed by selecting only the logs of the receivers whose proficiency needs to be assessed. Consequently, proficiency assessment can be performed even more efficiently.

[0076] In this embodiment, the determination device 5 further includes a second filtering unit 54. The second filtering unit 54 performs filtering to exclude logs relating to vehicles other than those that require operation by a vehicle operator. The data extraction unit 52 extracts system data from the logs after filtering by the second filtering unit 54.

[0077] This allows system data to be extracted while excluding logs for vehicles that do not require operator intervention (e.g., a normal ETC vehicle A1a that did not call an operator via intercom). Therefore, in the training data creation phase S1, system data can be extracted only from logs for the target vehicles (e.g., abnormal ETC vehicle A1b, non-ETC vehicle A2 that called an operator via intercom, and normal ETC vehicle A1a that called an operator via intercom). This improves the accuracy of the learning model created in the subsequent learning phase S2. As a result, the proficiency level of the toll collector can be determined with even greater accuracy.

[0078] Furthermore, in the judgment phase S3, system data can be extracted only from the logs of the vehicle being operated. Therefore, the proficiency level can be determined with even greater accuracy.

[0079] In this embodiment, the determination device 5 further includes a third filtering unit 55. The third filtering unit 55 removes logs when an inappropriate operation is performed by the receiver for determining proficiency. The data extraction unit 52 extracts system data from the logs after filtering by the third filtering unit 55.

[0080] This allows for the extraction of system data by excluding logs unsuitable for proficiency assessment (for example, logs when the wrong entry button is pressed, when the non-payment button is pressed, when the forced entry button is pressed, logs when there is a long time between vehicle entry and the toll collector being called via intercom, logs related to toll collection while equipment malfunctions are occurring, etc.). Therefore, in the training data creation phase S1, system data can be extracted only from logs suitable for proficiency assessment. This improves the accuracy of the learning model created in the subsequent learning phase S2. As a result, the proficiency level of toll collectors can be determined with even greater accuracy.

[0081] Furthermore, in the judgment phase S3, system data can be extracted only from logs suitable for proficiency assessment. Therefore, proficiency assessment can be performed with even greater accuracy.

[0082] In this embodiment, the proficiency acquisition unit 51 acquires proficiency information as a score. The proficiency determination unit 58 determines the proficiency of the recipient as a score.

[0083] This allows for a simplified assessment of proficiency levels. Furthermore, the assessment results can be easily understood. Therefore, the efficiency of proficiency assessment can be further improved.

[0084] (Hardware configuration) The determination device 5 of the above-described embodiment is implemented in the computer 1100 shown in Figure 9. Figure 9 is a schematic block diagram showing the configuration of the computer 1100 according to each embodiment. The computer 1100 includes a processor 1110, main memory 1120, storage 1130, and interface 1140.

[0085] The operation of each of the functions of the determination device 5 described above is stored in the storage 1130 in the form of a program. The processor 1110 reads the program from the storage 1130, loads it into the main memory 1120, and executes the above processing according to the program. The processor 1110 also allocates storage space in the main memory 1120 according to the program.

[0086] The program may be for the purpose of realizing some of the functions that the computer 1100 is to perform. For example, the program may perform functions in combination with other programs already stored in the storage 1130, or in combination with other programs implemented in other devices. In addition, the computer 1100 may be equipped with a custom LSI (Large Scale Integrated Circuit) such as a PLD (Programmable Logic Device) in addition to or instead of the above configuration. Examples of PLDs include PAL (Programmable Array Logic), GAL (Generic Array Logic), CPLD (Complex Programmable Logic Device), and FPGA (Field Programmable Gate Array). In this case, some or all of the functions realized by the processor 1110 may be realized by the integrated circuit.

[0087] Examples of storage 1130 include magnetic disks, magneto-optical disks, and semiconductor memory. Storage 1130 may be an internal medium directly connected to the bus of computer 1100, or an external medium connected to computer 1100 via interface 1140 or a communication line. Furthermore, if this program is distributed to computer 1100 via a communication line, computer 1100 that receives the program may expand it into main memory 1120 and execute the above processing. Storage 1130 may also be a tangible storage medium that is not temporary.

[0088] Furthermore, the program may be intended to implement some of the functions described above. In addition, the program may be a so-called differential file (differential program) that implements the functions described above in combination with other programs already stored in the storage 1130.

[0089] (Other embodiments) Although embodiments of this disclosure have been described in detail above with reference to the drawings, the specific configuration is not limited to these embodiments and may include design changes and the like that do not depart from the gist of this disclosure. In the above embodiment, lane L was assumed to be a mixed lane used by both ETC vehicle A1 and non-ETC vehicle A2, but it is not limited to this. Lane L may be an ETC-only lane equipped with an electronic toll collection system (ETC equipment 2), or it may be a non-ETC-only lane without ETC equipment 2.

[0090] In the above embodiment, the data extraction unit 52 extracts system data from logs recorded in the automatic toll collection machine 3 and logs recorded in the monitoring panel 4, but it is not limited to this. The data extraction unit 52 may extract system data from logs recorded in either the automatic toll collection machine 3 or the monitoring panel 4.

[0091] In the above embodiment, the data extraction unit 52 extracts system data after filtering by the first filtering unit 53, the second filtering unit 54, and the third filtering unit 55, but it is not limited to this. The data extraction unit 52 may extract system data after filtering by one or two of the first filtering unit 53, the second filtering unit 54, and the third filtering unit 55 has been performed. Furthermore, filtering by the first filtering unit 53, the second filtering unit 54, and the third filtering unit 55 does not have to be performed before data extraction by the data extraction unit 52. In addition, the order of filtering by the first filtering unit 53, the second filtering unit 54, and the third filtering unit 55 can be changed as appropriate.

[0092] In the above embodiment, the proficiency acquisition unit 51 acquires proficiency information as a score, and similarly, the proficiency determination unit 58 determines the proficiency of the payer as a score, but this is not limited to this. For example, the proficiency acquisition unit 51 may acquire proficiency information as characters or symbols, and similarly, the proficiency determination unit 58 may determine the proficiency of the payer as characters or symbols.

[0093] Figure 10 is a diagram showing the functional configuration of a determination device according to a modified example of the present disclosure. Figure 11 is a flowchart showing the procedure for the training data creation phase S1 of the determination method relating to a modified example of the present disclosure. Figure 12 is a flowchart showing the procedure for determination phase S3 of the determination method according to a modified example of the present disclosure. As shown in Figure 10, the determination device 5 may further include a log acquisition unit 59. The log acquisition unit 59 receives and acquires logs from either the automatic toll collection machine 3 or the monitoring panel 4.

[0094] In this case, as shown in Figure 11, in the training data creation phase S1, after acquiring proficiency information (step S11), the log acquisition unit 59 acquires logs from either the automatic fare collection machine 3 or the monitoring panel 4 (step S17). Filtering (steps S12-S14) and data extraction (step S15) are performed on the logs acquired in step S17.

[0095] Furthermore, as shown in Figure 12, in the determination phase S3, similar to the training data creation phase S1, the log acquisition unit 59 first acquires logs from either the automatic toll collection machine 3 or the monitoring panel 4 (step S36). Filtering (steps S31-S33) and data extraction (step S34) are performed on the logs acquired in step S36.

[0096] <Note> The determination device 5 and determination method described in each embodiment can be understood, for example, as follows.

[0097] (1) The determination device 5 according to the first embodiment includes: a proficiency acquisition unit 51 that acquires proficiency information evaluated for each work item of the proficiency of a toll collector who remotely operates an automatic toll collection machine 3 for collecting tolls from vehicles using a monitoring panel 4; a data extraction unit 52 that extracts system data related to the toll collector's operation for each work item from logs recorded in at least one of the automatic toll collection machine 3 and the monitoring panel 4; a teacher data creation unit 56 that creates teacher data from the system data and the proficiency information; a learning model creation unit 57 that creates a learning model having a correlation between the system data and the proficiency information from the teacher data; and a proficiency determination unit 58 that inputs the system data extracted from logs recorded in at least one of the automatic toll collection machine 3 and the monitoring panel 4, separately from the use for creating the teacher data, into the learning model and determines the proficiency of the toll collector for each work item.

[0098] As a result, once a learning model is created, the judgment device 5 can use the learning model to automatically determine the proficiency level of the collection officer for each work item.

[0099] (2) The determination device 5 according to the second embodiment is the determination device 5 according to the first embodiment, further comprising a first filtering unit 53 that filters out logs relating to recipients for whom proficiency level determination is not required based on the recipient's ID included in the log, and the data extraction unit 52 extracts the system data from the log after filtering by the first filtering unit 53.

[0100] This allows system data to be extracted while excluding logs related to highly skilled veteran collectors for whom proficiency assessment is unnecessary.

[0101] (3) The determination device 5 according to the third embodiment is the determination device 5 according to the first or second embodiment, further comprising a second filtering unit 54 that performs filtering to exclude the logs relating to vehicles other than the vehicles that are subject to operation and require operation by the receiver, and the data extraction unit 52 may extract the system data from the logs after filtering by the second filtering unit 54.

[0102] This allows system data to be extracted, excluding logs for vehicles that do not require operator intervention.

[0103] (4) The determination device 5 according to the fourth embodiment is a determination device 5 according to any one of the first to third embodiments, further comprising a third filtering unit 55 that excludes the log when an inappropriate operation for determining proficiency is performed by the recipient, and the data extraction unit 52 may extract the system data from the log after filtering by the third filtering unit 55.

[0104] This allows for the extraction of system data while excluding logs unsuitable for assessing proficiency.

[0105] (5) The determination device 5 according to the fifth embodiment is a determination device 5 according to any one of the first to fourth embodiments, wherein the proficiency acquisition unit 51 acquires the proficiency information in the form of points, and the proficiency determination unit 58 determines the proficiency of the recipient in the form of points.

[0106] This allows for a simplified assessment of proficiency levels. Furthermore, the results of the proficiency assessment can be easily understood.

[0107] (6) The determination method relating to the sixth aspect includes the steps of: acquiring proficiency information in which the proficiency of a toll collector who remotely operates an automatic toll collection machine 3 for collecting tolls from vehicles using a monitoring panel 4 is evaluated for each work item; extracting system data relating to the toll collector's operation for each work item from logs recorded in at least one of the automatic toll collection machine 3 and the monitoring panel 4; creating training data from the system data and the proficiency information; creating a learning model from the training data that has a correlation between the system data and the proficiency information; and inputting the system data extracted from logs recorded in at least one of the automatic toll collection machine 3 and the monitoring panel 4, separately from the use for creating the training data, into the learning model to determine the proficiency of the toll collector for each work item. [Explanation of Symbols]

[0108] 1…Toll collection system 2…ETC equipment 3…Automatic toll collection machine 4…Monitoring panel 5…Judgment device 21…Entry vehicle detector 22…Roadside antenna 23…Communication vehicle detector 24…Lane server 25…Start control unit 26…Exit vehicle detector 31…Camera 32…Operation reception unit 33…Toll collection processing unit 41…Display device 42…Input device 43…Instruction transmission unit 51…Proficiency acquisition unit 52…Data extraction unit 53…First filtering unit 54…Second filtering unit 55…Third filtering unit 56…Training data creation unit 57…Learning model creation unit 58…Proficiency determination unit 59…Log acquisition unit 1100…Computer 1110…Processor 1120…Main memory 1130…Storage 1140…Interface A1…ETC vehicle A1a…Normal ETC vehicle A1b…Abnormal ETC vehicle A2...Non-ETC vehicle α...Onboard equipment I...Island L...Lane R1...Communication area R2...Photography area X1...Entry detection position X2...Communication end position X3...Exit detection position

Claims

1. A proficiency acquisition unit acquires proficiency information that evaluates the proficiency of toll collectors who remotely operate an automated toll collection machine that collects tolls from vehicles using a monitoring panel, for each work item. A data extraction unit extracts system data related to the operator's operations for each work item from logs recorded in at least one of the automatic toll collection machine and the monitoring panel, A teacher data creation unit that creates teacher data from the aforementioned system data and proficiency information, A learning model creation unit creates a learning model that has a correlation between the system data and the proficiency information from the aforementioned training data, The learning model includes a proficiency determination unit that inputs the system data extracted from the logs recorded in at least one of the automatic fare collection machine and the monitoring panel, separately from the use for creating the training data, to determine the proficiency of the fare collector for each work item, A determination device equipped with the following features.

2. The system further includes a first filtering unit that performs filtering to exclude logs relating to recipients for whom proficiency assessment is unnecessary, based on the recipient's ID included in the log. The determination device according to claim 1, wherein the data extraction unit extracts the system data from the log after filtering by the first filtering unit.

3. The system further includes a second filtering unit that performs filtering to exclude logs relating to vehicles other than the vehicle being operated by the recipient, among the aforementioned vehicles. The determination device according to claim 1 or 2, wherein the data extraction unit extracts the system data from the log after filtering by the second filtering unit.

4. The system further includes a third filtering unit that removes the log when the recipient performs an inappropriate operation for determining proficiency level, The determination device according to claim 1 or 2, wherein the data extraction unit extracts the system data from the log after filtering by the third filtering unit.

5. The proficiency acquisition unit receives the proficiency information as a score, The proficiency determination unit determines the proficiency of the receipt recipient in terms of points, as described in claim 1 or 2.

6. A step of acquiring proficiency information, in which the proficiency of a toll collector who remotely operates an automatic toll collection machine for collecting tolls from vehicles using a monitoring panel is evaluated for each work item, The computer extracts system data related to the toll collector's operations for each work item from logs recorded in at least one of the toll collection machine and the monitoring panel. The steps include: The computer creating training data from the system data and the proficiency information; The steps include: the computer creating a learning model from the training data that has a correlation between the system data and the proficiency information; The computer inputs the system data extracted from the logs recorded in at least one of the automatic fare collection machine and the monitoring panel into the learning model, separately from the use for creating the training data, and determines the proficiency level of the fare collector for each work item. A determination method that includes this.

Images